The idealism of the early days of the internet in the mid 90s, perfectly captured in John Perry Barlow’s Declaration of the Independence of Cyberspace, gave way to concern and consternation as an increasingly lawless digital Wild West emerged over the coming years, with all the ills of social media and a wider attention economy being allowed to proliferate under the protection of Section 230. It took until the 2020s for governments to start legislating in earnest to try and tackle issues such as online safety, social media addiction and political disinformation. But are the regulatory measures actually solving the problems, or is a fundamentally new approach required?
Antisocial media
The focus of concern amongst those advocating for greater online regulation tends to be on the harms which arise from the use of social media platforms by children. The Molly Russell Foundation, set up by the father of 14 year old Molly Russell who took her own life after being exposed to thousands of images promoting suicide and self-harm on social media sites, aims to hold “government and regulators to account to make the online world safe for children and young people.” Across the pond, other bereaved parents and campaigners are mounting litigation against social media companies, with some degree of success.
Aside from the often deleterious and sometimes catastrophic effects of social media use on the mental health of children and young people, there is the scourge of misinformation propelled by AI slop used to create memes, and state sponsored disinformation campaigns designed to threaten political stability and democracy itself. Furthermore, the addictive nature of social media algorithms is leading to vast amounts of time being sucked up by the attention economy, potentially damaging productivity and economic growth.
Obviously not all the harms of the internet can be put at the door of social media channels. A wide array of dangers extend beyond the domains of Meta and TikTok, such as websites hosting pornographic and other age inappropriate material being easily accessed by children, personal bankruptcy caused by online gambling addiction, and the facilitation of organised crime via cryptocurrency and the dark web. The rapid rise of AI brings a whole array of new dangers such as deepfakes, chatbot addiction and associated AI induced psychosis, alongside serious threats to national security in the form of enhanced cyberattacks, the spectre of AI enabled bio-terrorism and democratic interference by hostile states. But even benign use of the internet can have negative impacts, with a recent study showing that the wellbeing of children with high exposure to “screens” can be harmed, with learning and language skills suffering.
The approach of government
Successive UK governments attempted to introduce age restrictions for pornography for about a decade, but this was repeatedly delayed. Regulation eventually arrived in the form of the Online Safety Act (OSA), which went far beyond mere age restrictions and attempted to tackle many different online harms to protect both children and adults. It converged with criminal law to outlaw new offences such as cyberflashing, and placed obligations on tech platforms to protect users from more traditional crimes such as fraud. Ofcom was tasked with enforcing the provisions of OSA but has struggled even with some of its more straightforward investigations, and has been criticised as lacking any teeth to effectively regulate the internet.
The UK has followed the example of Australia, which became the first country in the world to impose a ban on social media for under 16s. Many countries have now followed suit, with the UK recently announcing that the ban will take effect from March 2027. However, these bans have been criticised from many quarters, including the Molly Russell Foundation, pointing at their ineffectiveness, with children simply installing freely available Virtual Private Network (VPN) tools to circumvent geographical restrictions.
Commenting, Paul Bernal, Professor of Information Technology Law, UEA Law School, University of East Anglia, says:
“The original strategy of the Online Safety Act remains the primary approach. In essence, it means passing on responsibility for policing the net (and social media in particular) to the social networks themselves. Whilst it might seem attractive – the social media organisations have the resources and technical skills to do this where the government largely doesn’t – it is doomed to failure because the tech giants really don’t care, and do not take the government seriously, and Ofcom is a feeble regulator who has teeth but does not dare bite. Moreover, the government is still terrified of taking on the likes of Elon Musk.
“The twin ideas of age verification and social media bans for kids are doomed to failure and have disastrous side effects, taking away privacy from everyone, not just the kids. Kids will sidestep the regulations and start exploring the unregulated nether regions of the internet, which will be bad for everyone. The record from Australia is already terrible, and that’s just what we can see. Where kids are seriously avoiding being seen, the government won’t have any idea what they’re doing.”
According to William Richmond‑Coggan, a Partner at Freeths who specialises in data protection and technology law, the current approach is complicated by the lack of an agreed definition of harmful content and understanding of where the responsibility lies for mitigating harms:
“In the majority of cases, governments’ strategy towards online harms consists of shifting the responsibility: firstly onto the platforms where such harms are perhaps most likely to be encountered by the majority of users, making them responsible for establishing checks and safeguards, but then in turn shifting responsibility onto the users themselves, by prohibiting access to what is deemed to be harmful content. This reflects the twin challenges around this area, of firstly identifying what actually is meant by harmful content – particularly in the challenging area of lawful but harmful content – and secondly in actually policing compliance.”
What is the solution?
Since the current approaches to internet regulation seem relatively ineffective, what are the options for governments and regulators to make a real change? Paul argues that blocking entire social media networks should be on the table:
“What the government should be doing is regulating social media for everyone, not just for kids. They should be challenging the companies in a genuine way. Fining is not enough, and the threat of prosecuting senior executives, though included in the Online Safety Act, is an empty one, as the tech bros know very well that it won’t be used, except perhaps symbolically on a relative minion. Ofcom does not need new powers, it needs to use the real powers that it has.”
He argues that the Grok scandal was an ideal opportunity to act, saying they “could and should have blocked X in the UK entirely. Instead, they wrote empty letters and got empty concessions from X, not actually taking the problem away at all. Their blocking powers are pretty much the only ones that are meaningful, but they have to be used.”
The majority of social media moderation these days relies on algorithms and AI, so could this provide a solution? William says:
“Given the success that many online platforms have with identifying and acting against copyright infringements, it might be thought that there ought to be an algorithmic tool which could help with the identification of harmful material, but in numerous respects this is proving to be significantly more challenging. How is an algorithm to distinguish between a site which contains content promoting harmful eating or dietary practices, for example, from one which is intended to support young people who are trying to tackle their eating disorders. Content intended to assist individuals in identifying that they are subject to coercive or controlling behaviour and teaching them how to obtain help can be blocked by algorithms which are intended to protect against such behaviours. There are numerous reports of channels and feeds directed to women’s health issues being blocked or shut down because they are identified as containing ‘inappropriate’ content.”
He warns that another issue with reliance on an algorithmic method of regulation, is the necessity of using human inputs to train AI to make nuanced distinctions between harmful and helpful content “requiring individuals to be exposed to swathes of damaging material in the interests of producing a tool which is better at blocking it from others.” He further raises concerns “that the creation of such technologies will place considerable power in the hands of states which intend to curtail their citizens’ access to dissident material, creating algorithmically enhanced censorship and tracking tools which can prevent free expression and assist in the identification of dissidents.”
Besides emboldening Ofcom to act, or ramping up the ability of AI to remove harmful content, one idea to regulate the internet which has been proposed multiple times over the years is to introduce an “internet passport” such as in the form of digital ID. The theory is that this could automatically block access to age restricted material or other harmful content for children. But even if this is technically feasible and the hurdle of VPN circumvention can be surmounted, the dangers of government overreach and clampdown on freedom of speech, as alluded to by William above, would be significantly magnified.
What does the future hold?
Paul fears that governments are essentially in thrall to big tech and will therefore be unlikely to solve the issue of internet regulation any time soon:
“I have no optimism that the government (or any possible alternative government) will do anything other than continue along this useless path. They will try to convince themselves that it is working, and will say lots of nice words, but do nothing helpful. One thing they could and should do is leave X immediately – it would not be a regulatory measure, but it would have far more effect than any regulatory measure in sight. They won’t do it, though, just as they won’t use the powers that they have. The Tech giants have them by the short and curlies.”
William believes that a more proactive approach by regulators, alongside better education may provide the key to unlocking more effective online safety:
“In the UK, time is needed to see whether the existing online safety regime can be effective, but lessons need to be learnt from the data protection regime, and Ofcom needs to take an active approach to enforcement if the rules around online safety are to be taken seriously and respected. In parallel with that, better education is required. The plethora of online tools (such as VPNs but also generative AI capabilities) mean that young people and vulnerable adults are in practice still exposed to an extremely harmful online world. More needs to be done to ensure that people understand the scope and severity of the rules already in place, and to support users to make better and safer choices, before any further regulation is contemplated.”
The argument for education above regulation may eventually provide the solution to creating a safer online experience. If governments and regulators are unwilling or unable to properly protect children from online harms, it is necessary to ensure that parents are fully aware of the dangers of purchasing a smartphone for their kids and have the ability to easily manage parental controls. An effective education campaign would need to go beyond simple messaging, helping parents to stay updated on developing online safety threats such as AI and teaching them how to prevent circumvention through VPNs and fake accounts. A comprehensive online education programme would go beyond protecting children, providing guidance to adults on avoiding scams and harmful content, particularly for older generations who can struggle to navigate an ever more confusing and chaotic online world. An effective education campaign could even help to raise the awareness of citizens to disinformation perpetrated by foreign states and make people more resistant to social engineering. It will likely also be cheaper and achieve faster results compared to top-down regulatory efforts, avoiding costly protracted battles with Silicon Valley. So perhaps education will be the magic bullet of internet regulation, arming the sheriff of the digital Wild West.
Alex Heshmaty is technology editor for the Newsletter. He runs Legal Words, a human legal copywriting agency based in Bath. Email alex@legalwords.co.uk.
Photo by Diogo Monteiro on Unsplash.
The post Internet regulation: an insoluble problem? appeared first on Internet for Lawyers Newsletter.